Can Google understand your non-english page?

An observation...

[mipapage]

Google can search pages in other languages that are specific to your topic, no? Like if I put Widgets, it can return pages about widgets in languages other than what I input.

If so, something that just occured to me is this:

Google has a translate function that can translate pages to english (and other languages?). This must (might?) be tied to the technology that they use for the multi-language serps.

If it is, and you have a page in another language and you run it through the translator, I would imagine that the better the translation is makes, the better it will be at understanding your page, and therefore spitting it out in multi-language serps.

Just a thought. Not sure how signifigant this is...

[jon80]

I doubt if Google has such a powerful translator function.
Google still returns different Serps depending upon whether accented characters are used.

[heini]

>Google can search pages in other languages that are specific to your topic, no? Like if I put Widgets, it can return pages about widgets in languages other than what I input.

No, it can't. Only if widgets is somewhere in the page. Or in the links.

>the better the translation is makes, the better it will be at understanding your page, and therefore spitting it out in multi-language serps.

No. Ranking is totally independenat of translational capabilities. It's ranked for the exact word in the query.

All in all Google's capabilities of dealing with multi language content is rather limited.
Luckily that has no influence whatsoever on their ranking process.

[edited by: heini at 7:04 pm (utc) on May 7, 2003]

[mipapage]

I doubt if Google has such a powerful translator function.
Google still returns different Serps depending upon whether accented characters are used.

I'm using it right now, and it's good. Wrt your second point, I have three accented characters in my main keyphrase and get the same results with or without them.

[dididudu]

Correct me if I am wrong, I believe google just need to store the information in foreign languages in the same way as English. Now there is differences between unicode (2 byte = 1 character) and normal characters (Latin Linguistics). So, when you search "widget", you should never see "widget" in Chinese or Jap. shown up unless the English "widget" is used on the particular page. Just my 2 cents.

[mipapage]

Thanks Heini,

so much for that idea! (less to worry about anyway.)

[heini]

split off the discussion on accented characters [webmasterworld.com]

>you should never see "widget" in Chinese or Jap. shown up unless the English "widget" is used on the particular page
Yes, but a chinese page also turns up in results for a query on widgets, if widgets is in links pointing to the page.

[Chris_D]

It is my understanding that the internet treats each letter as a character. Forget meaning - different characters = different.

eg is e the same as é? No they aren't as far as computers are concerned.

e is alt0101 (hold down alt key, type 0101, then release alt key) whereas
é is alt0233

ie different.

Try a search (with your preferences set to all languages) on:
cafe - around 7.82 million
café - around 3.9 million
now set preferences to English only and you get
cafe - around 3.84 million
café - around 1.96 million

The word has to be on the page - or in the link text pointing to it. The word comprises characters. Computers know that e is alt0101 and é is alt0233 and ë is alt0235 and they are all different.

If you had a French page - in French - about a café - (and never had the word cafe on your site) but all the inbound links were from English speakers, with English language keyboards - and their link text was therefore cafe (because they didn't know HTML character coding) - you would also probably rank in searches for cafe.

Nothing about translations or understanding.

Hope this explains it.

Chris_D